U.S. patent number 4,965,576 [Application Number 07/290,500] was granted by the patent office on 1990-10-23 for error correcting system capable of effectively using an allowable frequency band.
This patent grant is currently assigned to NEC Corporation. Invention is credited to Seiichi Noda, Masayoshi Watanabe.
United States Patent |
4,965,576 |
Watanabe , et al. |
October 23, 1990 |
Error correcting system capable of effectively using an allowable
frequency band
Abstract
In an encoder for use in encoding each input signal unit of N
bits into each of error correcting codewords, each input signal
unit is grouped into a plurality of parts which have bit lengths N1
and N2 in consideration of a transmission rate on a transmission
path and which are individually encoded into partial encoded
codewords which individually include redundancy signals,
respectively. The partial encoded codewords may have different
error correction ability and different code lengths. The partial
encoded codewords are combined together into each of the error
correcting codeword which may be transmitted to the transmission
path at a transmission rate close to an allowable transmission rate
of the transmission path. A decoder decodes each error correcting
codeword into each of output signal units by dividing each error
correcting codeword into divided codewords and by individually
decoding the divided codewords.
Inventors: |
Watanabe; Masayoshi (Tokyo,
JP), Noda; Seiichi (Tokyo, JP) |
Assignee: |
NEC Corporation (Tokyo,
JP)
|
Family
ID: |
18205919 |
Appl.
No.: |
07/290,500 |
Filed: |
December 27, 1988 |
Foreign Application Priority Data
|
|
|
|
|
Dec 24, 1987 [JP] |
|
|
62-328047 |
|
Current U.S.
Class: |
341/94; 341/50;
714/746; 714/758 |
Current CPC
Class: |
H03M
13/05 (20130101) |
Current International
Class: |
H03M
13/00 (20060101); H03M 13/05 (20060101); H03M
013/00 () |
Field of
Search: |
;341/50,51,59,67,94
;371/32,38,40,51 ;360/15,48,53 |
References Cited
[Referenced By]
U.S. Patent Documents
Primary Examiner: Shoop, Jr.; William M.
Assistant Examiner: Young; Brian K.
Attorney, Agent or Firm: Sughrue, Mion, Zinn, Macpeak &
Seas
Claims
What is claimed is:
1. An encoder for encoding input information signal units into
error correcting codewords, respectively, each of which can correct
an error or errors on decoding each of the error correcting
codewords and which is transmitted from said encoder through a
transmission path having an allowable frequency band, said encoder
comprising:
grouping means responsive to said input information signal units
for grouping each input information signal unit into first through
M-th information signal subunits on the basis of said allowable
frequency band, where M is an integer greater than unity;
first through M-th encoding means coupled to said grouping means an
individually supplied with said first through M-th information
signal subunits, respectively, for individually encoding said first
through M-th information signal subunits into first through M-th
encoded signal units which include said first through M-th signal
information subunits and first through M-th redundancy signal
units, respectively; and
means for combining said first through M-th encoded signal subunits
into a single one of said error correcting codewords.
2. An encoder as claimed in claim 1, wherein each of said error
correcting codewords has a predetermined transmission frequency
band which is narrower than, and close to, said allowable frequency
band.
3. A decoder for decoding a sequence of reception signal units each
of which is produced by dividing each of input information signal
units into first through M-th information signal subunits on the
basis of an allowable frequency band of a transmission path through
which said reception signal units are received, by individually
encoding said first through M-th information signal subunits into
first through M-th encoded signal units which include said first
through M-th information signal subunits and first through M-th
redundancy signal units, respectively, and by combining said first
through M-th encoded signal units, where M is an integer greater
than unity, said decoder comprising:
dividing means supplied with each of said reception signal units
for dividing each of said reception signal units into first through
M-th divided codewords representative of said first through M-th
encoded signal units;
decoding means for decoding said first through M-th divided
codewords into first through M-th divided information signal
subunits representative of said first through M-th information
signal subunits; and
means for combining said first through M-th divided information
signal subunits into a reproduction of each of said input
information signal units.
4. An error correcting system comprising an encoder for encoding
input information signal units into error correcting codewords,
respectively, each of which can correct an error or errors on
decoding each of the error correcting codewords and is transmitted
through a transmission path having an allowable frequency band, and
a decoder supplied with said error correcting codewords as
reception signal units for decoding said reception signal unit into
each of output codewords that is a reproduction of each of said
input information signal units, said encoder comprising:
grouping means responsive to said each input information signal
unit for grouping said each input information signal unit into
first through M-th information signal subunits on the basis of said
allowable frequency band, where M is an integer greater than
unity;
first through M-th encoding means coupled to said grouping means
and individually supplied with said first through M-th information
signal subunits, respectively, for individually encoding said first
through M-th information signal subunits into first through M-th
encoded signal units which include said first through M-th signal
information subunits and first through M-th redundancy signal
units, respectively; and
means coupled to said encoding means for combining said first
through M-th encoded signal subunits into a single one of said
error correcting codewords;
said decoder comprising:
dividing means supplied with each of said reception signal units
for dividing each of said reception signal units into first through
M-th divided codewords representative of said first through M-th
encoded signal units;
decoding means for decoding said first through M-th divided
codewords into first through M-th divided information signal
subunits representative of said first through M-th information
signal subunits, respectively; and
means for combining said first through M-th divided information
signal subunits into each of said output codewords.
Description
BACKGROUND OF THE INVENTION
This invention relates to an error correcting system which
comprises an encoder and a decoder to transmit and receive a
sequence of error correcting codewords through a transmission path,
such as a radio channel.
In an error correcting system of the type described, error
correction has been performed in various manners so as to correct a
single error, a double error, a triple error, a t-tuple error, or
the like which might occur during transmission. At any rate,
redundancy bit or bits are added as a redundancy signal in an
encoder to each information signal unit (namely, input signal unit)
which is composed of a predetermined bit number of information
bits. A combination of the information bits and the redundancy bits
is sent as an error correcting codeword from the encoder to a
decoder. In the decoder, the error correcting codeword is decoded
into the information bits with error or errors corrected by the use
of the redundancy bits.
It is well known in the art that an increase of the redundancy bits
brings about an increase of bits which can be corrected by the
error correcting codewords. In other words, error correction is
given a high ability with an increase of the redundancy bits. The
ability of error correction may be referred to as an error
correction ability. However, an increase of the redundancy bits
results in a reduction of a transmission rate which is determined
by an allowable frequency band of a transmission path. Therefore,
the bit number of the redundancy bits should be selected in
consideration of the allowable frequency band of the transmission
path and the information bits included in each information signal
unit.
The bit number of the information bits discretely and extensively
increases as the number of the redundancy bits increases one by
one. Such a discrete increase of the information bits makes it
difficult to exactly adjust the error correcting codeword to the
allowable frequency band. Under the circumstances, the error
correcting codeword has a transmission rate which is very lower
than an allowable transmission rate allowed by the allowable
frequency band. This shows that an unused frequency band remains in
the allowable frequency band at the sacrifice of the error
correcting ability. In other words, neither the allowable frequency
nor the error correcting ability of the error correcting codeword
is effectively used in the conventional error correcting
system.
SUMMARY OF THE INVENTION
It is an object of this invention to provide an error correcting
system which can effectively use an allowable frequency band of a
transmission path.
It is another object of this invention to provide an error
correcting system of the type described, which can improve an error
correcting ability in each error correcting codeword.
According to an aspect of this invention, an encoder is for use in
encoding each of input signal units into each of error correcting
codewords. The encoder comprises grouping means responsive to each
of the input signal units for grouping each of the input signal
units into a plurality of signal parts or first through M-th
encoded signal units, encoding means coupled to the grouping means
for individually encoding, the signal parts into partial encoded
codewords or first through M-th encoded signal units which include
redundancy bits in addition to the signal parts, respectively, and
means for combining the partial encoded codewords into each of the
error correcting codewords.
According to another aspect of this invention, a decoder is for use
in decoding a sequence of reception signal units each of which is
produced by dividing each of the input signal units into a
plurality of signal parts, by individually encoding the signal
parts into partial encoded codewords which include redundancy bits
in addition to the input signal units, and by combining the partial
encoded codewords. The decoder comprises dividing means for
dividing each of the reception signal units into divided codewords
representative of the partial encoded codewords, decoding means for
decoding the divided codewords into divided information parts
representative of the signal parts, and means for combining the
divided information parts into a reproduction of each of the input
signal units.
BRIEF DESCRIPTION OF THE DRAWING
FIG. 1 is a signal format for use in describing a conventional
error correcting system;
FIG. 2 is a signal format for use in describing an error correcting
system according to a preferred embodiment of this invention;
FIG. 3 is a block diagram of an encoder used in the error
correcting system which can implement the signal format shown in
FIG. 2; and
FIG. 4 is a block diagram of a decoder which is for use in
combination with the encoder illustrated in FIG. 3.
DESCRIPTION OF THE PREFERRED EMBODIMENT
Referring to FIG. 1, description will be made as regards
conventional error correction for a better understanding of this
invention. In the example being illustrated, an error correcting
codeword of a code length B bits is composed of an information
signal unit of N bits and a redundancy signal of K bits which is
added to the information signal unit, where N and K are integers
which will presently be exemplified. Thus, the error correcting
codeword has a code length of (N+K) bits. The numbers N and K have
a relationship invariably determined at every code that may be a
linear code, such as a Hamming code, a parity check code, a BCH
code, or the like.
As to the Hamming code for correcting a single error, B and N are
given by:
From Equation (1), it is readily understood that N may be equal to
or smaller than 0, 1, 4, 11, 26, 57, 120, 247, 502 when K is equal
to 1, 2, 3, 4, 5, 6, 7, 8, and 9, respectively, as known in the
art. This shows that B becomes equal to or smaller than 1, 3, 7,
15, 31, 63, 127, 255, and 511 when K is changed from 1 to 9,
respectively.
As to the linear code for correcting a double error, N may be equal
to or smaller than 3, 11, and 20 when K is equal to 7, 9, and 10,
respectively. In other words, B is equal to or smaller than 10, 20,
and 30 when K=7, 9, and 10, respectively. The linear code for
correcting the double error has an error correction ability higher
than the Hamming code for the single error.
Under the circumstances, it is concluded that the code length B is
discretely changed with an increase of K and that the information
signal unit of N bits undesiredly becomes short with an increase of
the error correction ability.
In the meanwhile, the error correcting codeword is transmitted
through a transmission path, such as a radio channel, having a
predetermined allowable frequency band. However, such a discrete
change of the code length B makes it difficult to exactly adjust
the code length to the predetermined allowable frequency band.
Accordingly, an unused frequency band remains in the predetermined
allowable frequency band, as pointed out in the preamble of the
instant specification.
Referring to FIGS. 2 through 4, an error correcting system
according to an embodiment of this invention carries out error
correction by the use of a codeword which has a format illustrated
in FIG. 2 and which has a full or total code length B'. The
codeword is transmitted at a transmission rate through a
transmission path having an allowable frequency band, as described
in conjunction with FIG. 1. In FIG. 2, the full code length B' of
the codeword is not equal to the code length B of the codeword
illustrated in FIG. 1 but is selected in consideration of an
allowable transmission rate in the allowable frequency band of the
radio channel. It is assumed that the information signal unit has N
bits like in FIG. 1. In the example being illustrated, the
information signal unit of N bits is divided or grouped into a
first part of N1 bits and a second part of N2 bits. The first and
the second parts are individually followed by a first redundancy
signal of K1 bits and a second redundancy signal of K2 bits to form
first and second partial encoded codewords of B1 and B2 bits,
respectively. The first partial encoded codeword is combined with
the second partial encoded codeword to be transmitted as an error
correcting codeword of B' bits from the transmitter section 11 to
the receiver section 12 through the transmission path.
It is possible to adjust the transmission rate of the error
correcting codeword to the allowable transmission rate of the
allowable frequency band by selecting the bit numbers of the first
and the second parts. Therefore, the allowable frequency band can
be effectively used without an unused frequency band remaining in
the allowable frequency band. In addition, the first partial
encoded codeword of B1 bits may have an error correcting ability
different from that of the second partial encoded codeword of B2
bits. For example, the first partial encoded codeword of B1 bits
may have the error correction ability of a single error while the
second partial encoded codeword of B2 bits may have the error
correction ability of a double error.
On the other hand, it is assumed that the information signal unit
of N bits cannot be encoded into a codeword having the error
correction ability of a double error because such a codeword brings
about an excessive increase of a transmission rate which exceeds
the allowable transmission rate of the transmission path
illustrated in FIG. 1, although an unused frequency band is left in
the allowable frequency band.
When the codeword illustrated in FIG. 2 is compared with that
illustrated in FIG. 1, taking the above into consideration, it is
readily understood that the former has an excellent error
correction ability in comparison with the latter.
In order to perform the error correction by the use of the codeword
illustrated in FIG. 2, the error correcting system comprises a
transmitter section 11 (as shown in FIG. 3) and a receiver section
12 (as shown in FIG. 4) which is communicable with the transmitter
section 11 through the transmission path.
In FIG. 3, the transmitter section 11 comprises an encoder circuit
15, an encoder controller 16, a transmitter 17, and an antenna
(unnumbered). The encoder 15 is supplied from a signal source (not
shown) with a sequence of input signal units IN each of which is
composed of N bits like in FIGS. 1 and 2. In the illustrated
example, each of the input signal units IN is assumed to be given
at every input signal unit IN in a bit parallel fashion and is
delivered to a signal distributor 20 controlled by the encoder
controller 16, as will become clear. Each of the input signal units
IN is divided or grouped by the signal distributor 20 into the
first part of N1 bits and the second part of N2 bits, as mentioned
in conjunction with FIG. 2. The first and the second parts are
depicted at G1 and G2 in FIG. 3, respectively. The bit numbers N1
and N2 of the first and the second parts G1 and G2 are determined
in consideration of the error correcting ability and the allowable
frequency band (namely, the allowable transmission rate) of the
transmission path. Anyway, the signal distributor 20 serves to
group each of the input signal units IN and may be called a signal
grouping circuit.
Herein, it is assumed for brevity of description that the first
part G1 is encoded by a first encoder unit 21 into a first partial
codeword which has the error correction ability of a single error
while the second part G2 is encoded by a second encoder unit 22
into a second partial codeword which has the error correction
ability of a double error.
Division of each input signal unit IN is determined in an encoder
control circuit 23 of the encoder controller 16 in accordance with
a rule determined in relation to the allowable transmission rate.
In other words, the encoder control circuit 23 produces control
signals in accordance with the rule. Specifically, the encoder
control circuit 23 delivers a distribution control signal as one of
the control signals to the signal distributor 20 to make the same
distribute the first and the second parts G1 and G2 to the first
and the second encoder units 21 and 22, respectively. In this
event, the first and the second parts G1 and G2 are also
distributed or sent to first and second calculation circuits 26 and
27 of the encoder controller 16. The first and the second
calculation circuits 21 and 22 are controlled by calculation
control signals given as ones of the control signals from the
encoder control circuit 23.
Responsive to the calculation control signals, the first and the
second calculation circuits 26 and 27 are enabled to calculate
first and second redundancy bits in relation to the first and the
second parts G1 and G2 in a known manner, respectively. The first
and the second redundancy bits may be composed of K1 and K2 bits,
as shown in FIG. 2 and are delivered to the first and the second
encoder units 21 and 22. The first and the second redundancy bits
are added to the first and the second parts G1 and G2,
respectively. As a result, the first and the second parts G1 and G2
are encoded by the first and the second encoder units 21 and 22
into the first and the second partial encoded codewords which are
depicted at Pc1 and Pc2 and which have code lengths of B1 and B2.
The first and the second partial encoded codewords Pc1 and Pc2 are
combined together under control of the encoder control circuit 23
by a signal combiner 29 into a single one of the error correcting
codewords which is depicted at Cw in FIG. 3 and which is produced
in a bit series fashion with a format illustrated in FIG. 2. The
codeword Cw is sent from the signal combiner 29 through the
transmitter 17 to the antenna and is thereafter delivered as a
transmission signal through the radio channel to the receiver
section 12 shown in FIG. 4.
A combination of the first and the second encoder units 21 and 22
and the encoder controller 16 may be collectively referred to as an
encoder which serves to encode the first and the second parts G1
and G2 into the first and the second partial encoded codewords Pc1
and Pc2.
With this structure, the codeword Cw can be transmitted at the
transmission rate which is very close to the allowable transmission
rate of the allowable frequency band. Accordingly, it is possible
to effectively use the allowable frequency band.
The receiver section 12 illustrated in FIG. 4 comprises an antenna
(unnumbered), a receiver 30, a decoder circuit 31, a decoder
controller 32. The decoder controller 32 comprises a decoder
control circuit 33, a first calculator 36, and a second calculator
37. The decoder control circuit 33 produces decoder control signals
in accordance with a rule similar to that determined for the
encoder circuit 15 illustrated in FIG. 3. The first and the second
calculators 36 and 37 carry out calculations predetermined for
error correction in response to the control signals given from the
decoder control circuit 33. Each of the first and the second
calculators 36 and 37 is individually known in the art and
therefore will not be described in detail.
The transmission signal is received by the receiver 30 through the
antenna and is sent as a received codeword Cw' to the decoder 31.
The received codeword Cw' is identical with the error correcting
codeword Cw on occurrence of no error during the transmission of
the error correcting codeword Cw and is supplied to that
demultiplexer 40 of the decoder 31 which demultiplexes the received
codeword Cw into first and second demultiplexed codewords Pc1' and
Pc2' which are reproductions of the first and the second partial
encoded codewords Pc1 and Pc2, respectively, and which have code
lengths of B1 and B2 bits, as illustrated in FIG. 2. The first and
the second demultiplexed codewords Pc1' and Pc2' are delivered to
first and second decoder units 41 and 42 of the decoder 31 on one
hand and to the first and the second calculators 36 and 37 of the
decoder controller 32, respectively, on the other hand.
Supplied with the first and the second demultiplexed codewords Pc1'
and Pc2', the first and the second calculators 36 and 37 carry out
the calculations to detect presence or absence of errors in the
first and the second demultiplexed codewords Pc1' and Pc2', to
detect positions of the errors when any errors are present, and to
supply the first and the second decoder units 41 and 42 with error
position signals indicative of the error positions. More
specifically, the calculations carried out in the first and the
second calculators 36 and 37 may be divisions of the first and the
second demultiplexed signals Pc1' and Pc2' by polynomials
predetermined for the first and the second demultiplexed codewords
Pc1' and Pc2', respectively, as known in the art.
The first and the second decoder units 41 and 42 correct the errors
of the first and the second demultiplexed codewords Pc1' and Pc2'
in response to the error position signals, respectively.
Thereafter, the first and the second demultiplexed codewords Pc1'
and Pc2' which are subjected to error correction are decoded into
first and second decoded information signals G1' and G2' which are
reproductions of the first and the second parts G1 and G2 of FIG.
3, respectively. Therefore, the first and the second decoded
information signals G1' and G2' are composed of N1 bits and N2
bits, respectively. Thus, a combination of the demultiplexer 40,
the first and the second decoder units 41 and 42, and the decoder
controller 32 serves to decode the received codeword Cw into the
first and the second demultiplexed codewords Pc1' and Pc2' and may
be referred to as a decoder.
The first and the second information signals G1' and G2' are
combined by a combination circuit 44 into an output codeword OUT of
N bits which is a reproduction of the input codeword IN. The output
codeword OUT may be produced either in a bit serial fashion or in a
bit parallel fashion.
Practically, let data transmission be carried out by the use of the
error correcting system illustrated in FIGS. 2 through 4. In this
event, it is assumed for simplification of description that the
error correcting codeword Cw can be transmitted within the
allowable frequency band of the transmission path when a code
length of the codeword is equal to or shorter than 300 bits.
When an error correcting code of a single error is used in the
conventional error correcting system on the above-mentioned
assumption, the code length of B bits in a single one of the error
correcting codewords (FIG. 1) should be equal to or shorter than
255 bits. In other words, K is equal to 8 and N is equal to or less
than 247, as already described before. Thus, the allowable
frequency band affords to transmit an additional code of 45 bits.
However, such an additional code can not be transmitted in the
conventional error correcting system because B becomes equal to or
less than 511 when K becomes equal to 9 so as to increase N.
In the error correcting system according to this invention, it is
possible to transmit the input signal unit IN even when N is
increased, for example, to 267. For this purpose, the first and the
second parts G1 and G2 are assumed to be encoded into the first and
the second partial encoded codewords Pc1 and Pc2 which are for
correcting a single error and a double error, respectively. In this
case, the input signal unit IN of 267 bits is grouped into the
first part G1 of 247 bits and the second part G2 of 20 bits. This
shows that N1 and N2 are equal to 247 and 20, respectively. The
first and the second calculation circuits 26 and 27 calculate the
first and the second redundancy signals of 8 and 10 bits from the
first and the second parts G1 and G2, respectively. Therefore, K1
and K2 are equal to 8 and 10, respectively.
Subsequently, the first and the second redundancy signals are added
to the first and the second parts G1 and G2 to be encoded into the
first and the second partial encoded codewords Pc1 and Pc2 of 255
and 30 bits, respectively. Thus, it is readily understood that B1
and B2 (FIG. 2) are equal to 255 and 30 and that the total code
length B' of the single error correcting codeword Cw becomes equal
to 285. Therefore, the allowable frequency band is effectively used
in the error correcting system of this invention in comparison with
the conventional error correcting system. Moreover, the single
error correcting codeword Cw illustrated in FIG. 3 has the error
correction ability higher than that illustrated in FIG. 1 because
the codeword Cw partially includes the error correcting code which
is capable of correcting a double error.
Herein, supposing that a ratio of the total code length B to the
allowable transmission rate is defined as an allowable code ratio,
the ratio becomes equal to 285/300 in the error correcting system
according to this invention and 255/300 in the conventional error
correcting system. Therefore, the error correcting system according
to this invention can establish an improved allowable code
ratio.
While this invention has thus far been described in conjunction
with a preferred embodiment thereof, it will readily be possible
for those skilled in the art to put this invention into practice in
various other manners. For example, each input signal unit may be
divided into signal parts by a factor greater than two. In
addition, the signal parts may be encoded into various kinds of
partial encoded codewords which may be combined together with a
wide variety of code lengths.
* * * * *